VERA-MH Suicide Risk
Chatbots talk with simulated patients who show different levels of suicide risk. A judge grades the full conversation against a clinical rubric.
Spring Health ranks 28 models on suicide risk. Slingshot AI independently tested ten models using two patient simulators.
- Made by
- Spring Health and the VERA-MH Expert Council
- Released
- October 2025
- Kind
- Wellbeing and sycophancy
- Level
- Expert
- Scored
- Graded by a model
- Answers
- Open-ended
- Tasks
- All public
- Measures
- Propensity
- Behavior
- Misaligned with values
- Subject
- Clinical medicine, Psychology
- Language
- English
- Interaction
- Multi-turn
- Items from
- Generated
- Refresh
- One fixed set
- Run by
- Anyone
- In views
- Biology and medicine, Content safety
- Strongest model tested
- Claude Opus 5 #11 on the capability index
- Last model added
- August 2026
- Scores refreshed
- November 20th, 2026
GPT-5.6 Terra leads with 75.0 among models released by August 2026
Best score per model from the benchmark's own leaderboard, read on November 20th, 2026. Where a model ran with several settings or harnesses, we keep its best.
The newest model with a score here came out in August 2026, and nobody has tested a newer one. Among the 20 highest-ranked models on the capability index, these have no score yet:
- GPT-6.2 Astra
- Claude Fable 5.2
-
Gemini 4 Carbon
- GPT-6 Astra
- and 12 more
- 75.0 GPT-5.6 Terra
- 72.0 Claude Sonnet 5
- 72.0 GPT-5.6 Sol
- 71.0 GPT-5.6 Luna
- 68.0 Claude Opus 4.5
- 68.0 Claude Sonnet 4.6
- 68.0 GPT-5.5
- 68.0 Claude Opus 4.8
- 67.0 GPT-5.4
- 66.0 Claude Opus 4.6
- 64.0 Claude Opus 5
- 61.0 Claude Fable 5
- 60.0 GPT-5.2
- 60.0 Claude Opus 4.7
- 58.0 Claude Sonnet 4.5
- 42.0 Claude Sonnet 4
- 40.0
Gemini 3.7 Flash
- 39.0
Gemini 3 Flash
- 39.0
Gemini 3.6 Flash
- 38.0
Gemini 3.5 Flash-Lite
- 37.0 Grok 3
- 37.0
Gemini 3.5 Flash
- 35.0 Claude Opus 4
- 35.0
Gemini 2.5 Flash
- 33.0
Gemini 3.1 Pro
- 32.0 Grok 4
- 32.0
Gemini 3 Pro
- 28.0 GPT-4o
- 1 GPT-5.6 Terra 75.0
- 2 Claude Sonnet 5 72.0
- 3 GPT-5.6 Sol 72.0
- 4 GPT-5.6 Luna 71.0
- 5 Claude Opus 4.5 68.0
- 6 Claude Sonnet 4.6 68.0
- 7 GPT-5.5 68.0
- 8 Claude Opus 4.8 68.0
- 9 GPT-5.4 67.0
- 10 Claude Opus 4.6 66.0
- 11 Claude Opus 5 64.0
- 12 Claude Fable 5 61.0
- 13 GPT-5.2 60.0
- 14 Claude Opus 4.7 60.0
- 15 Claude Sonnet 4.5 58.0
- 16 Claude Sonnet 4 42.0
- 17
Gemini 3.7 Flash 40.0
- 18
Gemini 3 Flash 39.0
- 19
Gemini 3.6 Flash 39.0
- 20
Gemini 3.5 Flash-Lite 38.0
- 21 Grok 3 37.0
- 22
Gemini 3.5 Flash 37.0
- 23 Claude Opus 4 35.0
- 24
Gemini 2.5 Flash Sep 2025 35.0
- 25
Gemini 3.1 Pro 33.0
- 26 Grok 4 32.0
- 27
Gemini 3 Pro 32.0
- 28 GPT-4o Mar 2025 28.0
Source: Official VERA-MH Suicide Risk leaderboard
| # | Model | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | GPT-5.6 Terra | 75.0 | Jul 2026 | Official leaderboard | |
| 2 | Claude Sonnet 5 | 72.0 | Jun 2026 | Official leaderboard | |
| 3 | GPT-5.6 Sol | 72.0 | Jul 2026 | Official leaderboard | |
| 4 | GPT-5.6 Luna | 71.0 | Jul 2026 | Official leaderboard | |
| 5 | Claude Opus 4.5 | 68.0 | Nov 2025 | Official leaderboard | |
| 6 | Claude Sonnet 4.6 | 68.0 | Feb 2026 | Official leaderboard | |
| 7 | GPT-5.5 | 68.0 | Apr 2026 | Official leaderboard | |
| 8 | Claude Opus 4.8 | 68.0 | May 2026 | Official leaderboard | |
| 9 | GPT-5.4 | 67.0 | Mar 2026 | Official leaderboard | |
| 10 | Claude Opus 4.6 | 66.0 | Feb 2026 | Official leaderboard | |
| 11 | Claude Opus 5 | 64.0 | Jul 2026 | Official leaderboard | |
| 12 | Claude Fable 5 | 61.0 | Jun 2026 | Official leaderboard | |
| 13 | GPT-5.2 | 60.0 | Dec 2025 | Official leaderboard | |
| 14 | Claude Opus 4.7 | 60.0 | Apr 2026 | Official leaderboard | |
| 15 | Claude Sonnet 4.5 | 58.0 | Sep 2025 | Official leaderboard | |
| 16 | Claude Sonnet 4 | 42.0 | May 2025 | Official leaderboard | |
| 17 | | 40.0 | Aug 2026 | Official leaderboard | |
| 18 | | 39.0 | Dec 2025 | Official leaderboard | |
| 19 | | 39.0 | Jul 2026 | Official leaderboard | |
| 20 | | 38.0 | Jul 2026 | Official leaderboard | |
| 21 | Grok 3 | 37.0 | Apr 2025 | Official leaderboard | |
| 22 | | 37.0 | May 2026 | Official leaderboard | |
| 23 | Claude Opus 4 | 35.0 | May 2025 | Official leaderboard | |
| 24 | | 35.0 | Sep 2025 | Official leaderboard | |
| 25 | | 33.0 | Feb 2026 | Official leaderboard | |
| 26 | Grok 4 | 32.0 | Jul 2025 | Official leaderboard | |
| 27 | | 32.0 | Nov 2025 | Official leaderboard | |
| 28 | GPT-4o Mar 2025 | 28.0 | Mar 2025 | Official leaderboard |
Rankings by part
We read 5 parts from the benchmark’s own leaderboard, each ranking models on one part of the test, such as Detects potential risk, Confirms risk and Guides to human care. GPT-4o leads 1 of them, and 4 other models lead the rest. Arrows show how far a model moved from its overall rank among the models we track.
Detects potential risk
28 models we track have a score here.
- 1 GPT-4o Mar 2025 ↑27 97.0
- 2 Claude Sonnet 5 97.0
- 3
Gemini 3.7 Flash ↑14 97.0
- 4 Claude Opus 4 ↑19 96.0
- 5 Claude Sonnet 4.6 ↑1 96.0
- 6
Gemini 3.5 Flash ↑16 96.0
- 7 Claude Fable 5 ↑5 96.0
- 8
Gemini 3.6 Flash ↑11 96.0
- 9 Claude Sonnet 4.5 ↑6 95.0
- 10
Gemini 3 Pro ↑17 94.0
Confirms risk
28 models we track have a score here.
- 1 GPT-5.6 Sol ↑2 100.0
- 2 GPT-5.6 Luna ↑2 99.0
- 3 GPT-5.2 ↑10 95.0
- 4 GPT-5.4 ↑5 94.0
- 5 GPT-5.5 ↑2 91.0
- 6 GPT-5.6 Terra ↓5 91.0
- 7 Claude Sonnet 5 ↓5 82.0
- 8 Claude Opus 4.8 74.0
- 9 Claude Sonnet 4.6 ↓3 61.0
- 10 Claude Opus 4.5 ↓5 53.0
Guides to human care
28 models we track have a score here.
- 1 GPT-5.6 Terra 65.0
- 2 Claude Fable 5 ↑10 63.0
- 3 GPT-5.6 Luna ↑1 63.0
- 4 GPT-5.6 Sol ↓1 61.0
- 5 Claude Sonnet 4.5 ↑10 58.0
- 6 Claude Opus 5 ↑5 57.0
- 7 Claude Opus 4.6 ↑3 55.0
- 8 GPT-5.4 ↑1 53.0
- 9 GPT-5.5 ↓2 50.0
- 10 Claude Sonnet 5 ↓8 50.0
Supportive conversation
28 models we track have a score here.
- 1 Claude Sonnet 4.6 ↑5 99.0
- 2 Claude Opus 4.5 ↑3 98.0
- 3 Claude Opus 4.8 ↑5 92.0
- 4 Claude Opus 4.6 ↑6 91.0
- 5 Claude Sonnet 5 ↓3 91.0
- 6 Claude Opus 4.7 ↑8 89.0
- 7 GPT-5.6 Terra ↓6 84.0
- 8 Claude Opus 5 ↑3 78.0
- 9 Claude Fable 5 ↑3 75.0
- 10 GPT-5.6 Sol ↓7 71.0
Follows AI boundaries
28 models we track have a score here.
- 1 Claude Opus 4.6 ↑9 68.0
- 2 Claude Sonnet 4.6 ↑4 59.0
- 3
Gemini 3 Pro ↑24 58.0
- 4 Claude Opus 4.5 ↑1 54.0
- 5
Gemini 3.1 Pro ↑20 54.0
- 6
Gemini 2.5 Flash Sep 2025 ↑18 53.0
- 7 Claude Sonnet 4.5 ↑8 53.0
- 8
Gemini 3.5 Flash ↑14 53.0
- 9 GPT-4o Mar 2025 ↑19 50.0
- 10
Gemini 3.7 Flash ↑7 50.0
4 models have held number one on the official leaderboard since its release in October 2025
Lab claims count from the day the card came out, independent runs from the model’s release.
Source: Official VERA-MH Suicide Risk leaderboard
Scores over time
All the scores we have, one dot each, so a model tested at several settings appears more than once.
Source: Official VERA-MH Suicide Risk leaderboard
No lab model card reports it yet
We read the headline results of 540 launch posts from 57 labs, and none of them list this benchmark.
More in Safety and alignment
- FORTRESS In use Paired adversarial and benign national-security misuse prompts. Rubric-graded risk score on adversarial and over-refusal score on benign prompts. Claude Opus 5.5 5.78%
- StrongREJECT In use Forbidden-prompt jailbreak benchmark with a rubric grader that scores how useful a non-refusal actually is.
Inkling 98.6%
- Gray Swan IPI Arena benchmark In use Indirect prompt-injection attacks from public red-teaming competitions replayed on tool, coding and browser agents. Attack success rate at k attempts.
Gemini 4 Carbon 0%
- Covert misbehavior (Thinking Machines) In use Agentic tasks that check whether a model misbehaves and conceals it. Scored by the rate of covert misbehavior.
Inkling 2 Mini 8.30%
- Destructive actions (Thinking Machines) In use Agentic tasks that check whether a model takes destructive actions it was not authorized to take. Scored by the rate of destructive actions.
Inkling 2 Small 4.70%
- Misbehavior (Thinking Machines) In use Agentic tasks that check whether a model fabricates results, ignores instructions or hides what it did when a task is hard. Scored by the rate of misbehavior.
Inkling 2 Mini 9.80%
Data audit
All 9 checks pass. Every benchmark
Catalog record 2 pass
- Pass Task access All tasks are public.
- Pass Confidence Checked against a primary source.
Scores 3 pass 2 notes
- Pass Independent runs Scores from the official leaderboard (28 entries).
- Pass Models scored 28 models we track.
- Pass Newest score The newest is from August 2026, for Gemini 3.7 Flash.
- Note Frontier models None of the top 10 has a score. The strongest tested is Claude Opus 5 (#11).
- Note Headroom Scored as a plain number, with no fixed ceiling.
Leaderboard import 3 pass
- Pass Importer We read the board on vera-mh.com: 28 rows for models we track at the last read.
- Pass Last read November 20th, 2026.
- Pass Read worked The last read worked.
Consistency 1 pass 1 note
- Note Lab claims No model has both a lab number and an independent run.
- Pass Known issues No published audit found problems.